Historical stock data your backtests can trust
Every AgentTrading backtest runs on 20+ years of split- and dividend-adjusted daily data for US-listed stocks and ETFs, with the exact date range and data treatment printed on every result.
- 1 THESIS
- 2 EVIDENCE
- 3 BACKTEST
- 4 RISK
- 5 VERDICT
02 EVIDENCE · FUNDAMENTALS
04 RISK · IN PLAIN ENGLISH
Sample scenarios, not a live backtest of what you typed. Past performance does not guarantee future results. Educational analysis only, not financial advice.
In short
Historical stock data for backtesting needs to be split- and dividend-adjusted, because unadjusted prices quietly corrupt results: a 20-year SPY test that ignores reinvested dividends misses a large share of total return, and an unadjusted stock split looks like a crash. AgentTrading runs every backtest on 20+ years of split- and dividend-adjusted daily data for US-listed stocks and ETFs, with dividends reinvested and a 0.1% cost per trade assumed by default, and prints the exact date range and assumptions on every result so nothing is hidden in the data prep. You never download a CSV or clean a price series; you type the rule and the clean history is already under it. That is the practical gap between AgentTrading and asking ChatGPT, which cannot run a backtest on adjusted data and is known to invent prices, and it is the chore that form-driven tools like Portfolio Visualizer leave partly to you. Starter is $19 a month with 10 years of daily history and Analyst at $49 a month with the full 20+ year record, both billed monthly or yearly. Educational analysis only: clean data shows what happened, never what will happen.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
WHAT YOU GET - HISTORICAL DATA
Historical data, run on the bench
Split- and dividend-adjusted, always
Unadjusted prices make splits look like crashes and hide the dividend share of total return. Every series on the bench is adjusted before your rule ever sees it.
20+ years of daily bars
US-listed stocks and ETFs with dividends reinvested. Starter includes 10 years of daily history; Analyst and Pro run the full 20+ year record.
Zero data wrangling
No CSV downloads, no column cleaning, no survivorship spreadsheet. Type the rule; the clean history is already underneath it.
The range is printed on every result
Each backtest states its exact date range and data treatment, so two people can reproduce the same run and argue about the idea, not the data.
HOW IT WORKS - 4 STEPS
From a sentence to a stamped verdict
Pick a ticker or basket
US-listed stocks and ETFs, from SPY to single names. The adjusted daily series loads automatically.
State the rule
The data exists to answer questions. Describe the rule in plain English and confirm the restated card before it runs.
Run on the full record
The test covers up to 20+ years of adjusted daily data, dividends reinvested, 0.1% default cost per trade, all printed on the chart.
Reproduce it anytime
Because the range and treatment are logged, the same run gives the same answer next month, which is what makes it evidence instead of an anecdote.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
On the same bench
The data exists to be tested against: backtesting software is the front door and the trading strategy tester is where verdicts get stamped. Option chains are a separate dataset with separate economics, priced vendor by vendor on historical options data. The ChatGPT stock analysis comparison shows what happens without adjusted data, and ETF backtesting covers allocation tests on the same record.
QUESTIONS - ASKED AND ANSWERED
Historical data: the common questions
Where can I get historical stock data?
Data vendors, broker exports, and some free sites offer it, but for backtesting the data must be split- and dividend-adjusted or results quietly corrupt. AgentTrading runs every backtest on 20+ years of adjusted daily data for US-listed stocks and ETFs, so you never download a CSV or clean a price series yourself.
What is adjusted historical stock data?
Adjusted data corrects past prices for stock splits and pays back dividends as if reinvested, so the series reflects real total return. Without it, a split looks like a crash and a 20-year SPY test misses a large share of return. Every series on the AgentTrading bench is adjusted before your rule ever sees it.
How many years of historical data do you need?
Enough to cover more than one kind of market, ideally two decades, so a rule is not judged on a single bull run. AgentTrading runs up to 20+ years of daily history; Starter is $19 a month with 10 years and Analyst at $49 a month with the full 20+ year record.
Is free historical stock data reliable?
It can be usable, but free sources are often unadjusted, have gaps, or carry survivorship bias, and cleaning them is the chore that sinks most DIY backtests. AgentTrading provides adjusted daily data with the exact date range and treatment printed on every result, so two people can reproduce the same run.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
Your next idea deserves a verdict, not a hunch.
Bring a thesis or a ticker. AgentTrading restates the rule, shows the evidence, runs 20+ years of history, and stamps an honest verdict. You decide.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.